Nandikanti et al. (2026) Unearthing dominant controls of precipitation – streamflow interaction using information theory and explainable AI in large-sample hydrology

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-09-25
• Authors: Siva Sai Syam Nandikanti, Maheswaran Rathinasamy
• DOI: 10.1016/j.jhydrol.2026.136498

Research Groups

Department of Civil Engineering, Indian Institute of Technology Hyderabad, Kandi, Telangana, India
Department of Climate Change, Indian Institute of Technology Hyderabad, Kandi, Telangana, India

Short Summary

This study uses information theory and explainable AI to quantify the strength of precipitation-streamflow interaction across the Conterminous United States. The results reveal that P-Q interactions vary systematically across hydroclimatic settings and are governed by region-specific combinations of climatic, hydrologic, and physiographic factors.

Objective

• Investigate the dominant controls of precipitation-streamflow interaction using information theory and explainable AI in large-sample hydrology.

Study Configuration

• Spatial Scale: Continental (Conterminous United States)
• Temporal Scale: Daily

Methodology and Data

• Models used: Transfer Entropy (TE) and Shapley Additive Explanations (SHAP)
• Data sources: Daily precipitation, streamflow, and 38 other variables related to hydrologic, climatic, soil, and vegetation characteristics for 554 catchments

Main Results

• The spatial variability in TE is delineated into eight distinct regimes, with the lowest and highest P-Q interactions observed in mountainous snow-dominated regions and humid catchments, respectively.
• Runoff timing, flow signatures, vegetation dynamics, and snow fraction emerge as important controls in different regions.

Contributions

• This study provides a comprehensive understanding of the spatial heterogeneity in precipitation-streamflow interaction and the factors governing it.
• The findings demonstrate that P-Q interactions vary systematically across hydroclimatic settings and are governed by region-specific combinations of climatic, hydrologic, and physiographic factors.

Funding

• This research was funded by [project/program name] with reference code [reference code].

Citation

@article{Nandikanti2026Unearthing,
  author = {Nandikanti, Siva Sai Syam and Rathinasamy, Maheswaran},
  title = {Unearthing dominant controls of precipitation – streamflow interaction using information theory and explainable AI in large-sample hydrology},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136498},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136498}
}